Please note that CoLlAGe has been fully rebuilt as a Python module and is available as a PIP/Docker install. We are currently testing integrating the new version of CoLlAGe into 3D Slicer, as well as other platforms.
Please visit/star the NEW UPDATED repository at: https://github.com/radxtools/collageradiomics
This is an alpha implementation of the CoLlAGe radiomics descriptor built as a 3D Slicer plugin. An alpha version of a command-line executable is also available under Releases. Information on installing, building, and running the tool are available in the wiki.
CoLlAGe captures subtle anisotropic differences in disease pathologies by measuring entropy of co-occurrences of voxel-level gradient orientations on imaging computed within a local neighborhood. CoLlAGe is based on the hypothesis that disruption in tissue microarchitecture can be quantified on imaging by measuring the disorder in voxel-wise gradient orientations. CoLlAGe involves assigning every image voxel a ‘disorder value’ associated with the co-occurrence matrix of gradient orientations computed around every voxel. Details on extraction of CoLlAGe features are included in [1]. After feature extraction, the subsequent distribution or different statistics such as mean, median, variance etc can be computed and used in conjunction with a machine learning classifier to distinguish similar appearing pathologies. The feasibility of CoLlAGe in distinguishing cancer from treatment confounders/benign conditions and characterizing molecular subtypes of cancers has been demonstrated in the context of multiple challenging clinical problems.
[1] Prasanna, P., Tiwari, P., & Madabhushi, A. (2016). Co-occurrence of Local Anisotropic Gradient Orientations (CoLlAGe): A new radiomics descriptor. Scientific Reports, 6:37241.
This code comes under the following license: Apache 2.0 - Read file: LICENSE